That is words that is pronunced equally except the stress pattern differs. For example "digest /daɪˈdʒɛst/" (verb) and "digest /ˈdaɪdʒɛst/" (noun). Or even the case where the stress are of different kinds "anden /¹andɛn/" (swedish: the duck) and "anden /²andɛn/" (swedish: the spirit)

All the examples I can think of actually spells the same, but that is not a requirement (if you could find an example that doesn't). I don't know if there's any example where they are spelled differently.

  • I don't know the answer to your question. However, although the words 'absent and ab'sent may be spelled the same, they have different phonemes in (in other words they aren't pronounced the same way apart from their different stresses). In particular, the verb has an /e/ in the final syllable, whereas the adjective and preposition both have a schwa, /ə/. Commented May 7, 2023 at 11:34
  • 1
    @Araucaria-him damn schwas, I think I found a better example (in the list of such words "digest" is noted to use the same phonemes)
    – skyking
    Commented May 7, 2023 at 11:42
  • The Swedish example is a case of minimal pairs distinguished by tone. I don’t actually know if there is a term for those specifically, but they are cross-linguistically much more common than minimal pairs distinguished by stress (just consider largely monosyllabic and highly tonal Asian languages that will often have such minimal pairs for almost every syllable in the language). Commented May 7, 2023 at 13:47
  • 2
    Those are called homographs; spelled the same way, pronounced differently. [I think you mean the tonic accent is different.] If the tonic accent is different, the pronunciation changes. graduate [verb] and the person who gets the degree, noun.
    – Lambie
    Commented May 7, 2023 at 14:09
  • 3
    @Lambie The examples given in the question happen to be homographs, but stress-based minimal pairs need not be homographs (e.g., Spanish tomo ‘I take’ vs tomó ‘s/he took’), and homographs also include many pairs that aren’t stress-based (e.g., sow /soʊ/ ‘plant seeds’ vs sow /saʊ/ ‘female pig’). Commented May 7, 2023 at 18:41

1 Answer 1


The general term for two distinct words differing exclusively in terms of a single property is "minimal pair". There is no special technical term for "differing only in stress" (or nasality, or voicing), so we would construct a phrase like "stress minimal pair", "voicing minimal pair" and so on.

However, the entire basis for defining minimal pairs is radically different in the case of stress, as opposed to other properties. If we have [pat] versus [bat] or [pad] versus [pat], where "[pat]" is the phonetic output, we have minimal pairs for voicing because in the phonetic output there is exactly one segment in {[pat, bat]} which differentiates the two words. That is what it means to be "minimal". If instead we have {[pat, bɔt]}, the pair is non-minimal, even if under analysis it turns out that [ɔ] always derives from /a/ after a voiced obstruent. Instead, we have a co-dependency, or co-variation.

In [daɪˈdʒɛst] versus [ˈdaɪdʒɛst] you have co-variation where in the verb form, σ1 is unstressed and σ2 is stressed but in the noun σ1 is stressed and σ2 is unstressed (following your analysis / transcriptions – usually, the noun form is transcribed [ˈdaɪˌdʒɛst], with secondary stress on the second syllable, compared to [ˈbɪgɨst]). In that case, the difference is not minimal, there is co-variation.

Theoretically, for English, you would be looking for a pair of words where primary stress is on a particular syllable and the other syllables differ in having secondary stress versus no stress. Pairs like "latex" and "latest" are in that ballpark, except that there are numerous other differences, i.e. this is not minimal, it just illustrates that "1-2" and "1-0" are contrastive stress patterns. In English, vowel reduction messes up the quest for a true stress minimal pair between secondary stress vs. no stress. There simply are no stress minimal pairs, given that "minimal pair" is a way of classifying phonetic outputs.

Your invocation of spelling introduces some uncertainty as to what you are really asking, since it suggests a more specific interest in orthographic defects, where one cannot completely predict the pronunciation of words from their spelling (English has lots of that). Russian exemplifies this problem, that the spelling system does not tell you where the stress of a word is even though it is contrastive. Many languages have such orthographic defects for various properties, and stress differences are particularly omitted from spelling practices (e.g. Italian ancora 'anchor' vs 'more').

  • "The general term for two distinct words differing exclusively in terms of a single property is "minimal pair"." That is not true. They can have different tonic accents and be the same minimal pairs in terms of phones. They can be heteronyms or homophones.
    – Lambie
    Commented May 8, 2023 at 14:40
  • 1
    @Lambie Who's army? Commented May 10, 2023 at 13:57
  • @Araucaria-him A general term has nothing to do with generals. Are you making a joke?
    – Lambie
    Commented May 10, 2023 at 14:05
  • 1
    @Lambie I meant on whose authority do those claims stand? Commented May 10, 2023 at 14:06
  • @Araucaria-him What claims? Do you agree with the OP's assertion I quoted above? I certainly do not.
    – Lambie
    Commented May 10, 2023 at 14:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.